Yesterday the paper on std::elide, P3288R1, was discussed at the
British C++ committee meeting. Having given consideration to
contributions made by Gašper Ažman, Oliver Rosten, Lénárd Szolnoki and
Matthew Taylor, today I have made Revision No. 2:

    http://www.virjacode.com/papers/elide.htm

And here's the list of changes:

    R1 => R2
        * New section entitled 'Alternatives' to suggest that
std::optional and std::variant could be given an emplace_invoke
method.
        * std::elide can be used in a constant-evaluated context.
        * std::construct_at is given as an alternative to placement
new (note that the former can be used in a constant-evaluated
context).
        * The example of std::any is removed because T must be
copy-constructible.
        * The example of boost::static_vector is added.
        * The proposed wording for the standard clarifies that the
accessible typedef is called tag_elide
        * Correction of typo, is_elider -> has_tag_elide_true
        * Correction of typo, typename T -> typename... Params